October 3, 2005
By KEVIN CAMERON
New York TImes
PROUD new owners were once greeted with a universal request for some measure
of their car's performance.
"What'll she do?" curious neighbors would ask.
Crowded highways and police radar guns, along with electronic speed controls
built into engine computers, discourage today's drivers from exploring their
cars' limits. Instead, buyers try to gauge performance by comparing
horsepower ratings - usually without realizing that the numbers have not
necessarily been comparable.
A recent change in the procedures for measuring horsepower is leveling the
field, and causing the horsepower figures of some vehicles to be revised.
The rating of the Toyota Camry's 3-liter V-6 engine shrank nearly 10
percent - to 190 horsepower, from 210 - though the engine's specifications
and the car's performance are unchanged.
The power ratings of other models came in higher than expected: Cadillac
originally estimated the output of the supercharged V-8 in the 2006 STS-V
sport sedan would be 440 horsepower, but testing to the new standard
produced a rating of 469.
The idea of horsepower came from James Watt, the 18th-century inventor who
developed the unit of measure as a promotional device to equate the work
that could be done by his improved steam engines with the number of horses
they could replace.
The concept of horsepower is based on a force moving through a distance in a
given period of time. Watt decided to define one horsepower as the ability
to lift a load of 550 pounds at the rate of one foot every second.
Horsepower is measured by connecting the test engine to a dynamometer, a
machine designed to exert a steady, controlled resistance to the engine's
spinning crankshaft. Horsepower is computed from the measured twisting force
of the engine - its torque - and its rotational speed in revolutions per
The power of European autos is sometimes given in kilowatts; one hundred of
James Watt's horsepower are equivalent to about 75 kilowatts.
From an automaker's point of view, the bigger the horsepower number, the
more attractive the car will be to performance-conscious shoppers. In the
past, automakers logically interpreted test procedures to make their engines
seem more muscular.
Rating an engine too optimistically can land an automaker in trouble,
though. Last year, Hyundai spent tens of millions of dollars to settle a
class-action lawsuit with owners who claimed that the company had overstated
power, inflating the cars' value.
In some cases, automakers have chosen to understate horsepower - when
insurance companies threatened to raise rates for high-horsepower models,
Every industry has its standards organization, and for carmakers it is the
Society of Automotive Engineers, now 100 years old. The society's first
standards for horsepower measurement were published in 1917. In the 1970's,
power ratings advertised by carmakers changed from gross output to net
horsepower, reflecting a shift from testing bare engines to measuring them
with a full complement of accessories.
The society's new standard is known as J1349. David Lancaster, chairman of
the committee that developed the test rules, explained that new technologies
open new loopholes in test procedures, which must be periodically plugged.
Before 1971, for example, a manufacturer might connect the engine's exhaust
system to a vacuum pump, gaining power by eliminating back pressure, the
resistance to free flow caused by mufflers. Under the 2006 standard, a
complete production exhaust system, or one of equal back pressure, must be
What kinds of gray areas are corrected by the new code for testing engine
Many modern engine computers electronically adjust ignition timing, either
advancing it to take advantage of premium-grade gasoline or retarding it
slightly to keep the engine from knocking on regular fuel. Such systems use
special sensors to "listen" for knock, which can damage the engine.
A manufacturer could therefore accurately say its engine will run
satisfactorily on regular gas, yet perform the horsepower test on a
higher-octane blend, knowing it will automatically take advantage of that
fuel to produce slightly more power. Under the new standard, testing is done
with the minimum octane fuel required for that model.
Other loopholes have been closed, too. Power losses caused by friction
inside the engine can be reduced by the use of a thinner grade of oil or by
using hotter oil (oil flows more freely as its temperature rises); the new
testing standards require the oil to be the same grade specified for normal
service, at the temperature "exhibited in service with a fully warmed-up
Engines today drive an increasing number of accessories, including an
air-conditioning compressor and pumps for engine coolant and power-steering
fluid, all of which draw some power. Should they be operating when the
horsepower is tested? Should the electrical alternator be generating
electricity, and if so, how much? The revised standards spell out the
answers to questions like these - and they will surely be revised again as
new gray areas emerge.
Adoption of the new certified power standard, which includes a provision for
a third-party witness to the testing, is voluntary. General Motors says it
expects to apply it to all newly rated engines in the future; some makers
will use it for all existing engines.
Horsepower ratings may be impressive to consumers, but the numbers are often
meaningless in real-world traffic conditions. Specifications typically state
something like "220 horsepower at 6,000 r.p.m." but because this is peak
output, less power is produced at all other engine speeds.
In fact, most driving uses just a fraction of the engine's potential;
typically, cars need only about 25 horsepower to maintain highway speed on a
level road. Drivers never call on the full reserve of available horsepower
unless they hold the accelerator fully to the floor all the way to peak
Horsepower numbers are a coded message from the maker; a high number is like
bright sporty clothing, suggesting excitement. A moderate number parallels a
conservative suit, suggesting traditional virtues like reliability. Such
images are important in marketing.
A new performance measure that would take into account how people really
drive could be created. It would average the horsepower actually needed in
most daily driving.
The most direct way for drivers to determine the right amount of horsepower
for their driving needs requires no calculations at all - just a test-drive.
Yet another $.02 worth from a proud owner of a 1970 Mach 1 351C @